Your first recording

1. Download a speech model

Open Models → Speech models and download Whisper Large V3 Turbo. It runs in the browser and is the default for new meetings. The download is about 1.2 GB and happens once. See transcription models to compare it with Large V3.

2. Record

Pick whichever fits:

• An online meeting. Join it in Brave or Chrome as usual, open the Echo extension popup while the meeting tab is selected, tick Participants have consented to recording, and choose Start recording.
• An in-person meeting. In Echo choose New meeting → In person, allow the microphone, and record.
• An existing file. Choose Upload a recording on the overview and pick an audio file.

Recording covers each option in detail.

3. Transcribe

Open the meeting and choose Create transcript. With Settings → General → Transcribe after recording on, Echo starts on its own. Keep an Echo tab open while Turbo transcribes; it runs in that tab.

When it finishes you can play the audio, jump to any passage, rename speakers, edit text and save moments. Every edit creates a new version, so you can go back.

4. Notes (optional)

With Ollama installed and a model downloaded, open the meeting’s Notes tab and generate notes. Each note links back to the passages it came from, so you can check it against what was said.

5. Export or back up

Export a single meeting from its Export menu (Markdown, text, SRT, JSON or audio), or the whole library from Settings → Storage → Export library.